In a secondary analysis of the SPRINT trial enrolling 7,232 hypertensive adults without baseline atrial fibrillation, clinically significant sleep disturbance—defined as a patient-reported symptom score ≥2, updated repeatedly over time—was associated with a 78% higher adjusted risk of incident AF (HR 1.78; 95% CI 1.27–2.51). Each one-point increase in sleep score conferred a 26% greater AF risk. Reciprocally, among 6,079 participants free of baseline sleep disturbance, developing AF was associated with a 59% higher risk of subsequent clinically significant sleep disturbance (HR 1.59; 95% CI 1.16–2.16). Both associations held across prespecified subgroups with no significant effect modification, over a median follow-up of approximately 3.3 years yielding 189 AF events and 1,422 sleep-disturbance events.
What elevates this beyond prior work is the bidirectional, longitudinally updated design—most earlier studies assessed sleep at a single baseline snapshot and tracked only one direction of causality. The time-dependent Cox modeling here captures evolving exposure states, reducing misclassification bias. Mechanistically, sleep fragmentation is known to amplify sympathetic tone, inflammatory cytokines, and atrial ectopy—pathways that plausibly sustain AF once initiated, which could then further disrupt sleep architecture through nocturnal palpitations and anxiety.
Important caveats: SPRINT enrolled only hypertensive, non-diabetic adults with elevated cardiovascular risk, limiting generalizability. AF ascertainment relied on protocol ECGs rather than continuous monitoring, likely undercounting paroxysmal episodes. Sleep was patient-reported, not polysomnography-confirmed. As a preprint awaiting peer review, these findings should be interpreted cautiously—but the bidirectional signal is clinically actionable: treating sleep disturbance may warrant consideration as a modifiable AF risk factor.
